Predicted reliability and maintainability figures of merit are compared to values measured during demonstration tests and those experienced in the field. Five data processing, two radar, and four display consoles are considered. Results indicated that reliability predictions are generally conservative, possibly due to improvements in parts between the creation of part failure rate references and their use in new equipment. On the other hand, maintainability predictions were far better than field experience with the same equipment. Demonstration testing of maintainability, however, resulted in a lower test mean-time-to-repair (MTTR) than predicted, for each equipment considered, indicating a weakness in the present procedures of maintainability demonstration by simulation of failures.